APPROACHES TO ARABIC DIALECTS - Internet Archiveon many aspects of Arabic dialects, from negative constructions to folk tales, from diglossia to phonology. His didactic materials for

APPROACHES TO ARABIC DIALECTS

STUDIES IN SEMITICLANGUAGES AND LINGUISTICS

EDITED BY

T. MURAOKA AND C.H.M. VERSTEEGH

VOLUME XXXVIII



A Collection of Articles presented to Manfred Woidich on the Occasion of his Sixtieth Birthday

EDITED BY

MARTINE HAAK, RUDOLF DE JONG, KEES VERSTEEGH

BRILLLEIDEN • BOSTON

2004

This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

Approaches to Arabic dialects : A collection of articles presented to Manfred Woidich on theoccasion of his sixtieth birthday / edited by Martine Haak, Rudolf de Jong, Kees Versteegh.

p. cm. – (Studies in Semitic languages and linguistics ; v. 38)Includes bibliographical references and index.Consists of articles in English, French, and German.ISBN 90-04-13206-61. Arabic language–Dialects. I. Woidich, Manfred. II. Haak, Martine, 1959- III. Jong,

Rudolf Erik de. IV. Versteegh, C. H. M. V. Studies in Semitic languages and linguistics ;38

PJ6709.A75 2003492.7'7–dc21

2003050254

ISSN 0081-8461ISBN 90 04 13206 6

© Copyright 2004 by Koninklijke Brill NV, Leiden, The Netherlands

All rights reserved. No part of this publication may be reproduced, translated, stored ina retrieval system, or transmitted in any form or by any means, electronic,mechanical, photocopying, recording or otherwise, without prior written

permission from the publisher.

Authorization to photocopy items for internal or personaluse is granted by Koninklijke Brill provided that

the appropriate fees are paid directly to The CopyrightClearance Center, 222 Rosewood Drive, Suite 910

Danvers , MA 01923, USA.Fees are subject to change.

printed in the netherlands

v

CONTENTS

Preface ........................................................................................ viiBibliography Manfred Woidich ................................................ xi

The Arabic of Rabì 'a: A q6ltu dialect of Northwestern Iraq ............ 1Farida Abu Haidar

Zum Verbmodifikator là- in nordmarokkanischen Dialekten ................ 13Jordi Aguadé

Variability reproduced: A variationist view of the [µ]/[∂] opposition in modern Arabic dialects ................................................................ 21Enam Al-Wer

Die arabischen Dialekte von Jaffa und Umgebung ............................ 33Werner Arnold

Von an-"Àß6r (al-Qaßr) nach Ìgni (Ìgli): Ein Vorbericht zu einigen arabischen Dialekten der Provinz 6r-Ra“ìdìya (Marokko) ................ 47Peter Behnstedt

Les parlers arabes nomades et sédentaires du Fezzàn, d’après William et Philippe Marçais .................................................................... 67Dominique Caubet

Quadriliteral verbs in the Arabic dialects of eastern Arabia ................ 97Clive Holes

Technicalities and terminology of raiding and warfare as revealed by Bedouin oral narratives ................................................................ 117Bruce Ingham

Jüdisches, christliches und muslimisches Arabisch in Mosul ................ 135Otto Jastrow

Characteristics of Bedouin dialects in Southern Sinai: Preliminary observations ................................................................................ 151Rudolf de Jong

Variation and change in Arabic urban vernaculars ............................ 177Catherine Miller

Remarks on ideophones in Nigerian Arabic ........................................ 207Jonathan Owens

Negations in the dialect of es-Sal†, Jordan ........................................ 221Heikki Palva

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page v

Unmarked feminine nouns in modern Arabic dialects .......................... 237Stephan Procházka

Relative-clause marking in Arabic dialects: A preliminary survey ........ 263Jan Retsö

On verbal nouns in colloquial and literary Arabic ............................ 275Judith Rosenhouse

The Arabic dialect of women in Meknes (Morocco): Gender linked sound changes? ............................................................................ 291Harry Stroomer

De la grammaticalisation de ‘comme’ (comparatif ) en arabe .............. 309Catherine Taine-Cheikh

Deixis et focalisation: La particule ta en arabe de Yafi' (Yémen) ...... 329Martine Vanhove

Pidginization and creolization revisited: The case of Arabic ................ 343Kees Versteegh

Is there an Orientalist linguistics? .................................................... 359Abderrahim Youssi

Colloquial Arabic in the 17th century: Yùsuf al-Ma©ribì’s Egyptian- Arabic wordlist .......................................................................... 373Elisabeth Zack

Index of Varieties of Arabic .................................................... 391Index of Persons ........................................................................ 395

vi

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page vi

vii

PREFACE

To find colleagues and friends of Manfred Woidich willing to con-tribute to the collection of articles now in hand was not the mostarduous of tasks. When we first started to approach colleagues onthe subject of a Festschrift for Manfred—this was right under his nose,during the fourth AIDA conference in Marrakech in April of 2000—their reactions were heartwarming and encouraging. In part, thisenthusiasm to contribute must have been due to the scholarly stand-ing of Manfred Woidich. But perhaps even more so to his person-ality—modest, ambitious only to work hard and preferably in thebackground—which, almost as a matter of paradox, earned him acentral position in the field of Arabic dialectology as one of its lead-ing figures.

The bibliography of Manfred Woidich’s publications bears testi-mony to his achievements in the field. He has written extensivelyon many aspects of Arabic dialects, from negative constructions tofolk tales, from diglossia to phonology. His didactic materials for thestudy of Egyptian and Standard Arabic have become the preferredteaching manual in many language courses. The dialect atlas of theEgyptian dialects, which he published together with Peter Behnstedt,already stands as a monument of research in Arabic dialectology.Their joint introduction to dialect geography, due to appear in theHandbuch der Orientalistik series at Brill’s, will no doubt become essen-tial reading for anyone working in this field. The present collectionis an homage to his achievements.

As editors, we have deliberately chosen to include only articles ontopics in the field of Arabic dialectology rather than inviting allfriends and colleagues working in different fields. In the resultingvolume various approaches to Arabic dialectology are represented,reflecting the central themes of Manfred Woidich’s research. Thevalue of his work lies in the combination of his deep respect for finedetail—and the patience to strive for the full hundred percent whilecollecting field-work data—with a broad interest in languages in gen-eral, both on the theoretical and practical level. He keeps himselfcompletely with novel approaches in general linguistics, eager toapply new insights to as yet unanalysed phenomena; he is a true

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page vii

polyglot, in the sense of someone who loves language as well as languages.

Although unassuming in his professional activities, Manfred willalways take pride in hosting a dinner party at his home, servingpasta al dente with home-made pesto sauce, lavishly accompanied byexquisite Frankish wines—or German beer for those who prefer—all in the jovial atmosphere of Middle Eastern hospitality. All threeof us, in our different capacities, have had intensive contacts withManfred over the years: research, teaching, administration, the gen-eral state of affairs in Dutch Oriental studies and other topics, butthe highlights of these contacts have always been the dinner parties.

And yes, there is always a lot of talk about Arabic dialects. Henever misses an opportunity to point out an especially interestingArabic expression, usually from one of his beloved Egyptian dialects.With his encyclopaedic knowledge of these dialects he has acquiredquite a reputation as a latter-day Professor Higgins, even in Egypt.He is known to have asked complete strangers in the Nile Valley orthe Delta whether their mother came from some unknown villagein Upper Egypt—and he would usually ‘guess’ right, leaving hisanonymous ‘victim’ and those present in a state of utter bewilder-ment. He is a true master at finding a parallel between a word hehears in any context with a word or expression from some Egyptiandialect—a habit the Moroccan waiters in the restaurants he visitedduring the Marrakech conference became well aware of.

There is always a sense of finality to a list of publications in aFestschrift, which in this case would create quite the wrong impres-sion. We have therefore ventured to add an item ‘to appear’, theCairene Egyptian grammar, which in many respects will sum up hisformidable knowledge about the dialect of Cairo; it may well haveappeared by the time this book is presented to him. As we writethese lines, we are aware that the occasion of his birthday itself maynot be very pleasant for him: an entire afternoon wasted with fes-tivities, while he could have been at his computer, working on yetanother publication on the dialect of il-Ba“andi in the Dakhla oasis.We sincerely hope that he will forgive us for this intrusion when hehas had a chance to read the articles collected in this volume, andfind the festive afternoon memorable at least.

One of Manfred’s first experiences with dialect differences musthave been when his classmates in Germany would tease him withhis ach-Laut whenever he had come to the blackboard to be heard

viii

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page viii

by the teacher, calling “Woidich, freu dich! ” [voidix froi dix]. Today,on his sixtieth birthday, and now that he has been an inspiringteacher and a leading scholar himself for a respectable number ofyears, on behalf of all contributors to this volume, we should like tomake this our wish and add

kull6 sana w-inta †ayyib, ya Farìd!

Diemen—The Hague—Batenburg Martine HaakJanuary 2003 Rudolf de Jong

Kees Versteegh

ix

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page ix

This page intentionally left blank

BIBLIOGRAPHY MANFRED WOIDICH

1969 Negation und negative Sätze im Ägyptisch-Arabischen. Diss., UniversitätMünchen.

1973–1974 “Die 3.sg.f. Perfekt im Dialekt von il-Bi'ràt”. Mélanges de l’UniversitéSaint-Joseph 48.355–372.

1973a Review of Fatma M. Mahgoub, A Linguistic Study of CaireneProverbs, Bloomington & The Hague: Indiana University Press,1968. Zeitschrift der deutschen morgenländischen Gesellschaft 123.403–407.

1973b Review of Abdel Ghany A. Khalafallah, A Descriptive Grammarof Saeìdi Egyptian Colloquial Arabic, The Hague & Paris: Mouton,1969. Zeitschrift der deutschen morgenländischen Gesellschaft 123.401–403.

1974a “Ein arabischer Bauerndialekt aus dem südlichen Oberägypten”.Zeitschrift der deutschen morgenländischen Gesellschaft 124.42–58.

1974b Review of Motie Ibrahim Hassan, In-nàs wil-malik: People andKing, Folk Tales in the Cairene Dialect in Roman Transcription,Copenhagen: Copenhagen University Publications Fund, 1971.Zeitschrift der deutschen morgenländischen Gesellschaft 124.146–149.

1975 “Zur Funktion des aktiven Partizips im Kairenisch-Arabischen”.Zeitschrift der deutschen morgenländischen Gesellschaft 125.273–299.

1978 “Bemerkungen zu den arabischen Dialekten Mittelägyptens”.Zeitschrift für arabische Linguistik 1.54–63.

1979 “Zum Dialekt von il-'Awàm‰a in der östlichen ”arqiyya (Ägypten).I. Einleitung, grammatische Skizze und Volkskundliches”. Zeit-schrift für arabische Linguistik 2.76–99.

1980a “Das Ägyptisch-Arabische”. Handbuch der arabischen Dialekte, ed.by Wolfdietrich Fischer & Otto Jastrow, 207–242. Wiesbaden:O. Harrassowitz.

1980b “illi als Konjunktion im Kairenischen”. Studien aus Arabistik undSemitistik, Anton Spitaler zum siebzigsten Geburtstag von seinem Schülernüberreicht, ed. by Werner Diem & Stefan Wild, 224–238. Wies-baden: O. Harrassowitz.

1980c “Zum Dialekt von il-'Awàm‰a in der östlichen ”arqiyya (Ägypten).II. Texte und Glossar”. Zeitschrift für arabische Linguistik 4.31–60.

1980d (with Peter Behnstedt) “Zum Sprachatlas von Ägypten”. Zeitschriftfür arabische Linguistik 5.176–192.

1982a (with Peter Behnstedt) “Die ägyptischen Oasen: Ein dialekto-logischer Vorbericht”. Zeitschrift für arabische Linguistik 8.39–71.

1982b Review of Stig T. Rasmussen, Ma˙mùd Taymùr, il-Ma˙ba"‰aqam talattà“a‰: An annotated phonemic transcription, Copenhagen:Copenhagen University Publications Fund, 1979. Zeitschrift fürarabische Linguistik 7.91–92.

xi

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page xi

1983a Review of Mokhtar Ahmed, Lehrbuch des Ägyptisch-Arabischen,Wiesbaden: O. Harrassowitz, 1980. Zeitschrift für arabische Linguistik11.92–95.

1983b (with Peter Behnstedt). Karte Ägypten—Arabische Dialekte. (= TübingerAtlas des Vorderen Orients (TAVO) A VIII 12.) Wiesbaden: L. Reichert.

1984a “Volksmedizinisches aus den Oasen”. Papyrus 11/12 (Cairo).1984b Review of Charles Vial, L’égyptien, tel qu’on l’écrit: Glossaire établi d’après

un choix d’æuvres littéraires égyptiennes contemporaines, Cairo: InstitutFrançais d’Archéologie Orientale du Caire, 1983. Der Islam 61.355–356.

1985a (with Peter Behnstedt). Die ägyptisch-arabischen Dialekte. I. Einleitungund Anmerkungen zu den Karten. (= Beihefte zum Tübinger Atlas des VorderenOrients, Reihe B (Geisteswissenschaften), 50/1.) Wiesbaden: L. ReichertVerlag. 129 pp.

1985b (with Peter Behnstedt) Die ägyptisch-arabischen Dialekte. II. Dialektatlasvon Ägypten. (= Beihefte zum Tübinger Atlas des Vorderen Orients, Reihe B(Geisteswissenschaften), 50/2.) Wiesbaden: L. Reichert. 561 pp.

1985c Enige aspecten der hedendaagse tweetaligheid in Egypte. Inaugural lectureat the University of Amsterdam, May 20, 1985.

1985d Übungsbuch zur arabischen Schriftsprache der Gegenwart. Wiesbaden: L. Reichert. [Schlüssel 1986.]

1987a (with Peter Behnstedt). Die ägyptisch-arabischen Dialekte. I. Texte. Teil1, Nildelta. (= Beihefte zum Tübinger Atlas des Vorderen Orients, Reihe B(Geisteswissenschaften), 50/3.) Wiesbaden: L. Reichert. 311 pp.

1987b Review of Mohamed S. Anwar, Be and equational sentences in Egyptiancolloquial Arabic, Amsterdam: J. Benjamins, 1979. Mediterranean LanguageReview 3.132–134.

1988 (with Peter Behnstedt). Die ägyptisch-arabischen Dialekte. III. Texte. Teil2, Niltal und Oasen. (= Beihefte zum Tübinger Atlas des Vorderen Orients,Reihe B (Geisteswissenschaften), 50/3.) Wiesbaden: L. Reichert. 450 pp.

1989a “Langform versus Kurzform: Die Kardinalzahlwörter von 3 bis 10im Kairenischen”. Jerusalem Studies in Arabic and Islam 12.199–232.

1989b “illi ‘dass’, illi ‘weil’ und zayy illi ‘als, ob’: Zur Reinterpretation vonRelativsatzgefügen im Kairenischen”. Mediterranean Language Review4/5.109–128.

1989c Review of Renate Malina, Zum schriftlichen Gebrauch des KairinischenDialekts anhand ausgewählter Texte von Sa'daddìn Wahba, Berlin, 1987.Der Islam 66:1.169–171.

1989d “Zur Bildung der Verbalstämme in den Ägyptisch-Arabischen Dia-lekten: Der II. und der III. Stamm”. XXIII. Deutscher Orientalistentag,Würzburg (1985), Ausgewählte Vorträge, ed. by Einar von Schuler, 200–210. Stuttgart: F. Steiner.

1990a Ahlan wa Sahlan: Eine praktische Einführung in die Kairoer Umgangssprache.Wiesbaden: L. Reichert. [Schlüssel 1991.]

1990b “Einige Aspekte der Diglossie im heutigen Ägypten”. AmsterdamMiddle Eastern Studies, ed. by Manfred Woidich, 99–130. Wiesbaden:L. Reichert.

xii

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page xii

1990c Materialien zur Kenntnis des Kairenisch-Arabischen. (= Leermiddelen voorde studie van de Arabische Taal en Cultuur, uitgegeven door de VakgroepArabisch van de Universiteit van Amsterdam Nr. 2.) Amsterdam:IMNO. 280 pp.

1990d “/a/: Ein vernachlässigtes Kapitel aus der Morphophonologieeines arabischen Dialektes”. Amsterdam Middle Eastern Studies, ed.by Manfred Woidich, 131–159. Wiesbaden: L. Reichert.

1990e (ed.). Amsterdam Middle Eastern Studies. Wiesbaden: L. Reichert.215 pp.

1991a “Zur Entwicklung der Konjunktion a˙san-la˙san im Kairenischen”.Festgabe für Hans-Rudolf Singer zum 65. Geburtstag am 6. April 1990überreicht von seinen Freunden und Kollegen, I, ed. by Martin Forstner,175–193. Frankfurt am Main: P. Lang.

1991b “Die Formtypen des Zustandssatzes im Kairenischen”. Zeitschriftfür arabische Linguistik 23.66–98.

1991c “Short /a/ in Cairo Arabic morphophonology”. Semitic Studiesin Honor of Wolf Leslau on the Occasion of his Eighty-fifth Birthday,November 14th, 1991, II, ed. by Alan S. Kaye, 1632–1651.Wiesbaden: O. Harrassowitz.

1992a “Vorangestellte Demonstrativa im Kairenischen”. Jerusalem Studiesin Arabic and Islam 15.195–219.

1992b (with Jacob M. Landau). “The Baladiyyàt A˙mad ilFàr: A noteon a modern Egyptian manuscript text”. Manuscripts of the MiddleEast 6.59–70.

1993a “Die Dialekte der ägyptischen Oasen: Westliches oder östlichesArabisch?”. Zeitschrift für arabische Linguistik 25.340–359.

1993b (with Jacob Landau). Arabisches Volkstheater in Kairo im Jahre 1909:A˙mad ilFàr und seine Schwänke. (= Bibliotheca Islamica, 38.) Stuttgart:F. Steiner. 489 pp.

1993c Bibliographie zum Ägyptisch-Arabischen. Amsterdam: IMNO, Uni-versiteit van Amsterdam. [1st ed., 1989.]

1994a “Cairo Arabic and the Egyptian dialects”. Actes des premièresjournées internationales de dialectologie arabe de Paris, ed. by DominiqueCaubet & Martine Vanhove, 493–507. Paris: Institut Nationaldes Langues et Cultures Orientales.

1994b (with Peter Behnstedt). Die ägyptisch-arabischen Dialekte. IV. GlossarArabisch-Deutsch. (= Beihefte zum Tübinger Atlas des Vorderen Orients,Reihe B (Geisteswissenschaften), 50/4.) Wiesbaden: L. Reichert. 513 pp.

1995–1997 “Negation in the Egyptian Arabic dialect of the Dakhla Oasis:The morphological reinterpretation of a phonological rule”.Mediterranean Language Review 9.13–28.

1995a “Das Kairenische im 19. Jh.: Gedanken zum ˇan†àwì’s ‘Traitéde langue arabe vulgaire’”. Dialectologia Arabica: A Collection ofArticles in Honour of the Sixtieth Birthday of Professor H. Palva,271–287. (= Studia Orientalia, 75.) Helsinki: Finnish OrientalSociety.

xiii

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page xiii

1995b “Al-Sa'ìd: Dialects”. Encyclopaedia of Islam. New edition, VIII,866b–867b.

1995c (with Rabha Heinen-Nasr). Kullu Tamàm: Inleiding tot de Egyptischeomgangstaal. Amsterdam: Bulaaq. 322 pp. [2 ed., 1996; 3rd ed.1999; first appeared under the title of Zay6 ba'∂u: Inleiding totde Egyptische omgangstaal, Amsterdam: IMNO, 1993.]

1995d Review of Stephan Procházka, Präpositionen in den neuarabischenDialekten, Vienna, 1993. Orientalistische Literaturzeitung 90.58–65.

1996a “Some cases of grammaticalisation in Egyptian Arabic”. Proceedingsof the 2nd International Conference of l’Association internationale pourla dialectologie arabe, held at Trinity Hall in the University of Cambridge,10–14 September 1995, ed. by Joe Cremona, Clive Holes &Geoffrey Khan, 259–268. Cambridge: University PublicationsCentre.

1996b “Rural dialects of Egyptian Arabic: An overview”. CEDEJ:Egypte-Monde Arabe (Cairo) 27–28.325–354.

1997 “Upper Egyptian Arabic and dialect mixing in historical per-spective”. Humanism, Culture, and Language in the Near East: Studiesin honor of Georg Krotkoff, ed. by A. Afsaruddin & A.H.M. Zahniser,185–197. Winona Lake, IN: Eisenbrauns.

1998 “Aus den Erinnerungen eines Hundertjährigen: Ein Text imDialekt von Balà† in Ost-Dakhla/Ägypten”. Estudios de dialec-tología norteafricana y andalusí 3.7–33.

1999a (with Peter Behnstedt). Die ägyptisch-arabischen Dialekte. V. GlossarDeutsch-Arabisch. (= Beihefte zum Tübinger Atlas des Vorderen Orients,Reihe B (Geisteswissenschaften), 50/5.) Wiesbaden: L. Reichert.409 pp.

1999b Review of K. Mörth, Kardinalzahlwörter von ein bis zehn in denneuarabischen Dialekten, Vienna, 1997. Bulletin of the School of Orientaland African Studies 355–357.

2000 “The Arabic dialect of ilBa“andi at Dakhla oasis (Egypt)”.Proceedings of the Third International Conference of AIDA (AssociationInternatonale de Dialectologie Arabe) held in Malta, 29 March–2 April1998, ed. by Manwel Mifsud, 145–150. Malta: Salesian Press.

2002 “Zum Dialekt von al-Qaßr in der Oase Dakhla (Ägypten)”.‘Sprich doch mit deinen Knechten Aramäisch, wir verstehen es!’: 60.Beiträge zur Semitistik. Festschrift für Otto Jastrow zum 60. Geburtstag,ed. by Werner Arnold & Hartmut Bobzin, 821–840. Wiesbaden:O. Harrassowitz.

2003 (with Jan Hoogland & Kees Versteegh). Woordenboek Nederlands-Arabisch/Arabisch-Nederlands. 2 vols. Amsterdam: Bulaaq.

To appear Grammatik des Kairenischen. (= Porta Linguarum Orientalium, 000.).Wiesbaden: O. Harrassowitz.

xiv

HAAK/DE JONG_F1_v-xiv 9/1/03 11:48 AM Page xiv

1

THE ARABIC OF RABÌ'A: A Q6LTU DIALECT OF NORTH-WESTERN IRAQ

Farida Abu HaidarLondon

1. Introduction

“Not too long ago, in the fifties and sixties, our knowledge of EgyptianArabic was practically limited to one dialect, i.e. the dialect of Cairo.It was considered the Egyptian Arabic dialect per se” (Woidich1994:493). This statement can apply to several other regions in theArab world where the dialect of the capital city is sometimes takento be representative of the speech of the country as a whole. In fact,for a long time, whenever anyone spoke of Iraqi Arabic they usuallymeant the dialect of Baghdad. Manfred Woidich has been instru-mental in bringing to our notice varieties of Egyptian Arabic, besidesthe dialect of Cairo. Similarly, Otto Jastrow (1978), covering anothercorner of the Middle East where Arabic is the principal language,has considerably broadened our knowledge of Iraqi Arabic by pro-viding descriptions of some of the major q6ltu dialects of the regionand identifying several hitherto unknown ones. Jastrow collected valu-able data in situ long before major population movement to and fromthe q6ltu-speaking region of northern Iraq led to dialect loss anddialect shift, particularly in small towns and villages.

Jastrow (1994:121–122) has divided the q6ltu dialects of northernIraq and Anatolia into six main groups: Mardin, Siirt, Diyarbakir,Kurdistan, Tigris, and Euphrates. He has also subdivided the Tigrisdialects into three branches: Mosul, Tikrit, and the non-Muslimdialects of central and southern Iraq. In this study I wish to surveysome points of phonology and morphology in the spoken Arabic ofRabì'a, a dialect of the Mosul branch of the Tigris group.1

1 The present study is the first part of a project on the spoken Arabic of theregion of Jabal Sinjar in Iraq.

HAAK/DE JONG_F2_1-12 9/1/03 11:48 AM Page 1

2

Rabì'a is a small town in the province of Mosul in Iraq, just insidethe border with Syria. It is situated at approximately 100 km north-west of the city of Mosul on the railway line connecting Mosul withAleppo.2 Rabì'a, up until the 1990s, was a fairly thriving town whosepredominantly Muslim, Arabic-speaking inhabitants maintained closecontact with neighbouring towns in Syria, and especially Tall Kushik,only 5 km from Rabì'a, and the nearest town to it. After the GulfWar and the introduction of border patrols, lack of freedom of move-ment and constant political upheavals in northern Iraq forced mostof the young and able-bodied to leave Rabì'a.3 Many fled to south-ern Turkey where they found refuge among the Arabic-speakingcommunities of Anatolia. Others made the journey to western Europeas asylum seekers.

The original inhabitants of Rabì'a are sedentary, and their livelihoodhas depended for generations on agriculture and sheep-rearing. Overthe years the sedentary population was joined by people of bedouinstock, mostly of ”ammari origin, who moved north in search of fer-tile pastures for their flocks, and settled in Rabì'a and neighbouringvillages. Differences between those who are of sedentary origin andothers of bedouin provenance are reflected in their speech as wellas in their religious observance. Those of bedouin origin are Shiiteand speak g6l6t 4 Arabic, while the sedentary inhabitants of the townitself are mostly Sunni, their speech being a variety of q6ltu Arabic.

The present study deals only with Rabì'a q6ltu Arabic (RA), q6ltuvarieties being the speech of a large number of the Arabic-speakingMuslim and non-Muslim inhabitants of the province of Mosul.5 Myfindings are based on five hours of recordings of the speech of sevenone-time inhabitants of Rabì'a, as well as on direct one-to-one inter-views, carried out between October 2000 and August 2001. The twowomen and five men, on whose speech this study is based, left Iraq

2 Right up until the Iran-Iraq war, trains along this line used to start their jour-ney in Baghdad, their final destination being Istanbul.

3 Because of large-scale emigration and lack of up-to-date records, it is impossi-ble to know the number of inhabitants that still live in Rabì'a.

4 Blanc (1964:30), who coined the terms g6l6t and q6ltu, transcribed them as geletand qeltu, where e stands for IPA 6. Some scholars, however, prefer to transcribethem as gilit and qiltu.

5 Johnstone (1975:89) states that q6ltu dialects “until fairly recently were spokenat least as far south as Baghdad”. One should add that for a long time the Arabicspoken in the province of Mosul consisted only of q6ltu dialects, spoken along withvarieties of Neo-Aramaic.


ì' 3

between 1995 and 1998. The two oldest are a married couple liv-ing in Turkey. The husband was born in 1962 and his wife in 1963.Among the other five are a younger married couple, born in 1971and 1973, now living in Switzerland. The remaining three men, bornin 1974, 1978 and 1980, are at present in Germany. Four of the menhad been shepherds in Iraq. The fifth had worked on his family’ssmall holding where they planted mainly wheat and barley. The twowomen were housewives. Only two of the men and the older womanhad had elementary education, ranging from two to three years.6

2. Phonology

2.1 Consonants

Like all q6ltu dialects, RA is characterised by the retention of uvu-lar plosive q and velar plosive k, allophones of g6l6t g and ‘ respec-tively. Thus compare:

RA Baghdadq: qarày6b garày6b ‘relatives’

halqadd halgadd ‘so much’ts6wwaq tsawwag ‘to buy provisions, he bought provisions’x6rqa x6rga ‘rag’rqàq rgàg ‘thin, flaky bread’ (a speciality of northern Iraq)

k: klèb ‘làb ‘dogs’b6ki b6‘a ‘to cry’˙akayt ˙6‘èt ‘I spoke’f6kk fa‘‘ ‘jaw’samak s6ma‘ ‘fish’

g and ‘, however, are realized in some loanwords in RA, e.g., g6llàb6yya‘tunic’ (< Egyptian Arabic, where g is the reflex of MSA [ ). g canalso be found in loans from Turkish where k of the donor languagebecomes g in RA,7 as, for example:

6 In dialect studies in situ I invariably rely on data from informants belonging todifferent generations. However, as I was unable to go to Rabì'a in person, mychoice of informants was limited to those I could find outside Iraq, who all hap-pened to be under the age of forty. Because of population movement after the GulfWar it has become increasingly difficult to study dialects in situ in some of the moreinaccessible regions of Iraq.

7 The following loanwords occur also in the Arabic of Mosul and Baghdad.


RA Turkishgòmlag gömlek ‘vest’yalag yelek ‘waistcoat, sleeveless cardigan’rang renk ‘colour’

k becomes g before b in the form akbar ‘bigger, larger’, realized inRA as agbar.8

‘ occurs in loanwords which include the following: ‘ày ‘tea’, ‘m6ntu‘cement’ and qa‘a© ‘smuggled, illegal’.

In a number of g6l6t and q6ltu dialects p can occur in some forms,borrowed either from Turkish or European languages. Phoneme p,however, does not occur as frequently in RA as it does in the Arabicof Baghdad for example. An example of a form with p in RA ispaßapòr† ‘passport’. The following are a few examples where p hasbeen preserved in the dialects of Baghdad and Mosul, but not incomparable RA forms:

RA Mosul Baghdadban†arùn pan†a©òn pan†arùn ‘trousers’b6rdàyi pa©dàyi parda ‘curtain’bìkàb pìkàp pìkàp ‘pick-up truck’

In a few borrowings from Western languages b of the donor lan-guages is realized as p in g6l6t varieties and in the Mosul dialect also.The two most frequently occurring examples of b > p are pàß ‘bus’and pàys6k6l ‘bicycle’. In RA these forms are realized as bàß andbàys6k6l respectively.

Interdentals are usually realized as dentals in RA, with ∆ > t, ≈ > d and µ > ∂:9

Mosul RAm6∆6l m6t6l ‘like’∆nèn tnèn ‘two’hà≈a hàda ‘this’≈ìb dìb ‘wolf ’µa©abtùnu ∂arabtùnu ‘I hit him’˙aµµ ˙a∂∂ ‘luck, fate’

8 This is the only example of k > g in my data.9 Dentals for interdentals is a characteristic of the speech of Christian Iraqis.

Jastrow (1994:120) found this phenomenon also “in most Jewish dialects of IraqiKurdistan, in the dialect of the Jews of Nusaybin/Qàm6“li . . . and in the Diyarbakirgroup in Anatolia”. Jastrow (1994), moreover, points out that there are in the q6ltu-speaking region of Southern Turkey, Northern Iraq and Northern Syria as manyas three different sound shifts involving interdentals.

4


There were no instances of ∆ > s, ≈ > z or µ > Ω in the idiolectssurveyed here.

As far as MSA ∂ and Ω are concerned, most Iraqi speech vari-eties do not distinguish between the two. In g6l6t varieties, and insome q6ltu dialects, notably Mosul, the two MSA phonemes have thesame sound and are realized as a continuant, whereas in ChristianBaghdadi Arabic, for example, as well as in RA, they are realizedas a stop.10 This complete loss of phonemic distinction between MSA∂ and Ω is not characteristic of Iraqi dialects alone, but obtains alsoin some other varieties of Arabic (Beeston 1970:18, n. 1).

The r > © shift, a salient feature of the Tigris group as a whole( Jastrow 1994:120), does not occur in RA. Thus compare:

Mosul RA©6˙tu r6˙tu ‘I went’©6bi rabi ‘he grew up’a©nab arnab rabbit’ka©©àt karràt ‘times, instances’†à© †àr ‘it flew’˙a[èy6© ˙a[èy6r ‘stones, pieces of rock’

An interesting phenomenon in RA is the shift n > l in the form©alam ‘sheep’.11 Elsewhere in the region the MSA form ©anam hasbeen preserved.

As in a number of dialects of the Tigris group, s and ß in RAbecome z and Ω in the contiguity of b, as in the forms zbù' ‘week’( g6l6t sbù' ), zbèy6' ‘weeks’ and Ωbà'i ‘my finger’ ( g6l6t 6ßba'i ). s and ßare retained where they are separated from b by a vowel, as in asèbì'‘weeks’, the more formal plural of zbù', and aßàb6' ‘fingers’.

The alveolar affricate [ is realized as ≥ by some informants, dueperhaps to Syrian influence. The following forms were heard withboth [ and its Levantine allophone ≥:

[6bb and ≥6bb (pl. [bèb/≥bèb) ‘well, watering-hole’˙à[èy6r and ˙a≥èy6r ‘pieces of rock’ßà[ and ßà≥ ‘griddle (for baking bread)’

10 Transcribed as µ and ∂ respectively.11 Procházka (2002) has found that the n > l shift occurs in the Arabic of south-

ern Turkey. He gives as an example Turkish dönüm > dlùm (pl. dlùmàt) ‘a unit ofmeasurement’.

ì' 5


MSA © in ©asal ‘to wash’ is realized as x in RA, thus: xasìl ‘washing’,xasal6t ‘she washed’, etc. x < © occurs also in the Arabic of Baghdad,but not in the dialect of Mosul where MSA © is preserved.

The pharyngeal fricative ' is sometimes elided and followed by along vowel, as in the following examples:

mòr6f ‘I don’t know (< mà a'r6f or ma a'r6f )’bàx6d à†i ‘I take and give (< bàx6d a'†i )’ya'm6l mà wè˙6d ‘he works with someone’ (< ya'm6l ma'a wè˙6d ).

In this last example the first ' is not elided.As in a number of q6ltu dialects, h is elided in the form 6“h6r

‘months’ when it is preceded by the cardinal numbers 3 to 10, e.g.tlat 6“6r ‘three months’, s6tt 6“6r ‘six months’, etc. In g6l6t dialects h isusually retained.

2.2 Vowels

2.2.1 Short vowelsThe short vowels in RA are a and 6 in medial position and a, i andu in final position.

a in medial position occurs in stressed and unstressed open sylla-bles: waraqàyi ‘piece of paper’, bala˙a ‘date, palm-tree’, manàra ‘minaret’,ba'6d ‘after’, ˙awìs ‘clothes’, labas6t ‘she wore’.

a occurs in closed syllables in a guttural or emphatic environment,or where it is in the contiguity of r and a back vowel: ma'[ùn ‘(tomato)paste’, karra ‘once’, ballù†a ‘acorn’, ˙ayyi ‘snake’, ˙abbi ‘grain’, ba††a‘duck’, “qadd ‘how much?’.

a occurs in final position where the preceding consonant is a gut-tural, an emphatic or r: karra ‘once’, x6rqa ‘rag’, s6yyàra ‘car’, ballù†a‘acorn’, bala˙a ‘date, palm-tree’, †aq†aqa ‘noise’.

6 occurs in closed syllables: ˙6lwi ‘pretty [fem.sg.]’, w6˙di ‘one[fem.sg.]’, n6[màyi ‘star’, [6bb ‘well, watering-hole’, [6wwa ‘below, under-neath’, m6n[al ‘scythe’, s6yyàra ‘car’, n6zzèli r6˙˙àli ‘people constantlyon the move [lit. people who stay the night and then move on]’,qab6l ‘before’, ba'6d ‘after’, ˙ar6b ‘war’.

There are, however, in my data two forms with medial a where6 might have been expected. These are madrasi ‘school’ and fatti ‘brothwith pieces of bread’. This could be because both forms are loan-words, the first from MSA and the second, a culinary speciality,from outside Rabì'a.

6


In vCCaC forms referring to colour, the initial vowel is 6 in non-emphatic, non-guttural environments. Thus compare: 6swad ‘black’,but a˙mar ‘red’, ax∂ar ‘green’, aΩraq ‘blue’, abya∂ ‘white’.

i occurs in word-final position following non-gutturals and non-emphatics: sani ‘year’, n6[màyi ‘star’, ˙6lwi ‘pretty [fem.sg.]’, w6˙di ‘one[fem.sg.]’, xàli ‘maternal aunt’, qaryi ‘village’.

u occurs in 1st. ps. sg. perfect verbs: q6ltu ‘I said’, “6ftu ‘I saw’,sama'tu ‘I heard’, nasaytu ‘I forgot’.

u occurs also as a suffixed 3rd ps. sg. pronoun: sama'nànu ‘we heardit’, 'aßaru ‘he squeezed it’, abùnu ‘his father’, baytu ‘his house’.

In the q6ltu dialects of the Tigris group the final vowel of the fem-inine morpheme is not elided in the construct state. In g6l6t dialects,on the other hand, the comparable vowel is invariably elided. Thuscompare:

q6ltu-Iraqi Baghdad Mosul RAq6ßßa qußß†a q6ßß6tu q6ßß6tu ‘his story/tale’˙adìqa ˙adìqta ˙adìq6tu ˙adìq6tu ‘his garden’s6yyàra s6yyàrathum s6yyà©6t6m s6yyàr6t6n ‘their car’

Similarly, on the affixation of pronominal suffixes to the form màl‘of, belonging to’ an anaptyctic vowel is added in RA and otherTigris branch q6ltu dialects. No such vowel occurs in g6l6t Arabicwhen the pronoun is in the singular. Thus compare:

Baghdad RAmàlti màl6ti ‘mine’màltak màl6tak ‘yours [masc.sg.]’màlt6‘ màl6tki ‘yours [fem.sg.]’màlta màl6tu ‘his’

But màlatna màl6tna ‘ours’màlatkum màl6tk6n ‘yours [comm.pl.]’màlathum màl6t6n ‘theirs’

According to Johnstone (1975:91), “In very few dialects can thereoccur a series of short open syllables”, and “in the dialect of Tikritthe vowel of the post-initial syllable is elided”. Johnstone gives amonghis examples katbat ‘she wrote’ and waqfu ‘they stood’. In RA, how-ever, the short vowel of the second syllable is not elided, so that theabove two examples are usually katab6t ‘she wrote’ and waqafu ‘theystood’. Compare also:

ì' 7


8

Baghdad RAs6m'aw sama'u ‘they heard’˙6lbat ˙alab6t ‘she milked’l6bsat labas6t ‘she got dressed’

Johnstone (1975:92) states that in Tikrit Arabic “there are no formsin which a final consonant cluster occurs”. One should point outhere, that in most Iraqi Arabic dialects, be they of the q6ltu or g6l6tgroup, an anaptyctic vowel occurs invariably between the two finalconsonants of comparable MSA CvCC forms. Thus:

MSA Baghdad RAbard bar6d bar6d ‘cold’ba'd ba'ad ba'6d ‘after’†abl †abul †ab6l ‘drum’˙uzn ˙6z6n ˙az6n ‘sorrow’

2.2.2 Long vowelsThe long vowels in RA are è, ì, à, ù, and occasionally ò, all occur-ring in medial position.

à and è: A salient feature of all q6ltu dialects is the "imàla (à > è )that distinguishes this group of dialects from g6l6t Iraqi Arabic. "Imàlais a feature that is widespread in RA. Thus compare:

Baghdad RAnà'6m nè'6m ‘soft’nsà'6d nsè'6d ‘we help’wà˙6d wè˙6d ‘one’fawàt6˙ fawèt6˙ ‘wakes [pl. of fàt6˙a]’˙amàya ˙amèyi ‘my brother-in-law’

"Imàla does not occur in post-guttural, post-emphatic environmentsor in the contiguity of r with a back vowel. Thus: qàl ‘to say, hesaid’, xàli ‘my maternal uncle’, nahàr ‘day’, naßàra ‘Christians’, ˙a††àbìn‘wood gatherers’.

ì: Vowel ì occurs mainly in adjectives of the CaCìC type and plu-rals of the CaCèCìC type, e.g. CaCìC: kabìr ‘big’, katìr ‘numerous’,ba'ìd ‘far’, qarìb ‘near’, xafìf ‘light’, ˙awìs ‘clothes’.12

12 This last form is unusual. In the Mosul dialect and the Arabic of Christiansof Baghdad and Basra the word for ‘clothes’ is ˙wès.


CaC√CìC: dakèkìn ‘shops’, ˙awèwìn ‘animals’, xawèrìf ‘sheep’.ù: RA ù, as in abùyi ‘my father’, axùyi ‘my brother’, aqùl ‘I say’,

a“ùf ‘I see’, is sometimes the equivalent of ò in comparable forms inthe Mosul dialect. Thus compare:

Mosul RAzbò' zbù' ‘week’tannò© tannùr ‘earthenware oven [for baking bread]’ballò† ballù† ‘acorn’ma'[òn ma'[ùn ‘(tomato) paste’malµòm mal∂ùm ‘dried okra threaded with string’

ò: Long ò is rare, but can occur sometimes in loanwords, e.g., gòm-lak ‘vest’, b6z-zòr ‘by force’. I also found in my data two compoundforms with medial ò, mòr6f ‘I don’t know (< mà a'r6f )’ and lòx6r ‘theother (< MSA al-àxar)’. Long medial ò can also be heard in a“lòn‘how?’ although this forms seems to be in free variation with kayf.13

2.2.3 DiphthongsDiphthongs aw and ay occur in RA where ò and è might be expected.Thus compare:

Mosul RAmòß6l mawß6l ‘Mosul’mòt mawt ‘death’dò“ak daw“ak ‘mattress’hònìki hawnik ‘there’y6[òn y6[awn ‘they come’bèt bayt ‘house’“èx “ayx ‘venerable old man’baqètu baqaytu ‘I stayed’˙akèna ˙akayna ‘we spoke, we talked’

According to Jastrow (1994:120) “the diphthongs ay and aw havebeen preserved in many q6ltu dialects”.

13 The use of kayf is no doubt due to contact with Syrian Arabic where Syriankîf is adapted to the sound system of RA. See following paragraph on diphthongs.

ì' 9


10

3. Morphology

3.1 Verbs

As in all q6ltu dialects, the 1st ps. sg. of the perfect in RA ends in-tu, as, for example, r6˙tu ‘I went’, sama'tu ‘I heard’, t'6llamtu ‘I learnt’,ts6wwaqtu ‘I bought provisions’, rabaytu ‘I grew up’, ˙alabtu ‘I milked’,x6ttartu ‘I made yoghurt’, [ara“tu ‘I ground [lentils, cracked wheat,etc.]’, 'a[antu ‘I kneaded [bread]’.

The imperfect in RA is frequently preceded by a modifier ka refer-ring to a continuous past action, or by prefix b- denoting a habitualaction:

ka y6[awn ‘they used to come’ka t6bß6†u ‘she used to spread it’ka t6˙l6b 6n-na'[èt ‘she used to milk the ewes’ka y6sqùh6n ‘they used to water them’bàx6d à†i ‘I give and take [lit. I take and give]’baqùm w aq'6d ‘I move about [lit. I get up and sit down]’btàk6l t6“rab w tnèm ‘she eats, drinks and sleeps’

Of the two modifiers ka is the one that occurs more frequently.Although it functions as a particle, ka is in all likelihood derivedfrom the auxiliary verb kàn. Prefix b- is less common, and could bedue to Syrian influence (cf. Behnstedt 1997; Arnold 1998). Whenusing more than one imperfect in the same sentence, informantstended to use b- with the first verbal form only, as can be seen fromthe last three examples.

3.2 Nouns

Like some q6ltu dialects, notably Christian Iraqi Arabic and the dialectof Mosul, a number of feminine unit nouns have an ending in -àyi,e.g., waraqàyi ‘a piece of paper’, n6[màyi ‘a star’, k6bbàyi ‘a patty ofcrushed wheat and minced lamb’,14 bay∂àyi ‘an egg’.

Some feminine nouns have an ending in -6yyi. There are twoexamples in my data: g6llàb6yyi ‘cloak’ and †annùr6yyi ‘skirt’.

14 This form is not to be confused with the Syro-Lebanese k6bbàyi ‘(drinking)glass’. It should be pointed out that k6bbi (kubba in g6l6t dialects) is the regional spe-ciality of the province of Mosul, its main ingredient being b6r©6l ‘crushed wheat’,the staple diet of the region.


ì' 11

The frequently occurring plural forms are of the CaC√CìC andCaC√Ci type, where the second syllable is always long. CaC√CìC:dakèkìn ‘shops’ (sg. d6kkèn), ˙awèwìn ‘animals’ (sg. ˙aywèn), xawèrìf ‘sheep’,15

masèkìn ‘poor people’ (sg. m6skìn), malè'ìn ‘scoundrels’ (sg. mal'ùn).CaC√Ci: n6[èmi ‘stars’, †alèmi ‘(thick) loaves (of local bread)’, asèmi‘names’, faqàri ‘poor (pl.)’, [a˙à“i ‘asses’.

3.3 Bound morpheme -6n

Jastrow (1994:120) states that there is “a very stable isogloss” in q6ltuArabic which separates the Anatolian branch from other q6ltu branches,and that is the m/n ending of pronominal morphemes. Jastrow foundthat in the Anatolian branch forms like bèthum (Dèr iz-Zòr), bayt6m('Aqra) ‘their house’ and [ìtum (Dèr iz-Zòr), [ìt6m ('Aqra) ‘you [comm.pl.]came’, for example, are bayt6n and [ìt6n respectively in Mardin. Thisfeature seems to be characteristic of RA also, as can be seen fromthe following examples:

Mosul RAy6sqùh6m y6sqùh6n ‘they water them, they give them to drink’˙alab6t6m ˙alab6t6n ‘she milked them’bèt6m bayt6n ‘their house’màl6t6m màl6t6n ‘theirs’b6nt6m b6nt6n ‘their daughter’

4. Conclusion

As can be seen from the above examples the Arabic spoken by thesedentary inhabitants of Rabì'a is a variety of the Tigris branch ofq6ltu Arabic. RA has a number of similarities with the Arabic of thecity of Mosul, one of the better known dialects in the region. [Forthis dialect see Jastrow’s contribution to this volume, Eds.] WhereRA differs considerably from the dialect of Mosul at the phonologicallevel, however, is primarily in the retention of MSA r, the allophoneof Mosul Arabic ©. At the morphological level, the main difference

15 I do not have the singular of xawèrîf in my data. Informants constantly referredto ‘sheep’ as ©alam (sg. and pl.) or ©alamèt (pl.). For the singular they seemed to usena'[i ‘ewe’.


between RA and Mosul Arabic is the pronominal morpheme end-ing in n instead of m. Modifier ka referring to a continuous pastaction does not occur in Mosul where the auxiliary verb kèn (< MSAkàna) is used instead. As far as verbal prefix b- is concerned, it isdifficult to ascertain from the idiolects surveyed here whether this isa characteristic feature of RA or merely a loan from neighbouringSyrian Arabic. Only a thorough investigation of a cross-section ofRA speakers of different ages and of both sexes would provide amore in-depth knowledge of the dialect as a whole.

Bibliographical References

Arnold, Werner. 1998. Die arabischen Dialekte Antiochiens. Wiesbaden: O. Harrassowitz.Beeston, A.F.L. 1970. The Arabic Language Today. London: Hutchinson University

Library.Behnstedt, Peter. 1997. Sprachatlas von Syrien. Wiesbaden: O. Harrassowitz.Jastrow, Otto. 1978. Die mesopotamisch-arabischen q6ltu-Dialekte. 2 vols. Wiesbaden:

F. Steiner.—— 1989. Der arabische Dialekt der Juden von 'Aqra und Arbìl. Wiesbaden: O. Harrassowitz.—— 1994. “The q6ltu dialects of Mesopotamian Arabic”. Actas del Congreso Internacional

sobre Interferencias Lingüísticas Arabo-Romances y Paralelos Extra-Iberos, ed. by JordiAguadé, Federico Corriente & Marina Marugán, 119–123. Zaragoza.

Johnstone, Thomas Muir. 1975. “The spoken Arabic of Tikrît”. Annual of the LeedsUniversity Oriental Society 7.89–109.

Procházka, Stephan. 2002. Die arabischen Dialekte der Çukurova (Südtürkei ). Wiesbaden:O. Harrassowitz.

Woidich, Manfred. 1994. “Cairo Arabic and the Egyptian dialects”. Actes des pre-mières journées internationales de dialectologie arabe de Paris, ed. by Dominique Caubet& Martine Vanhove, 493–507. Paris: INALCO.

12


13

ZUM VERBMODIFIKATOR LÀ-IN NORDMAROKKANISCHEN DIALEKTEN

Jordi AguadéUniversidad de Cádiz

Bekanntlich gebrauchen fast alle marokkanischen Dialekte einenVerbmodifikator, um die allgemeine oder aktuelle Gegenwart auszu-drücken.1 Am meisten verbreitet sind kà- und tà-, wobei kà- so ziem-lich überall zu finden ist (Fes, Rabat, Casablanca, Tanger, Tetouan,Larache usw.),2 inzwischen auch bei jüngeren Sprechern mancherDialekte, die traditionell andere Verbmodifikatoren kannten. Dagegenwird tà- vor allem in den südlichen Dialekten verwendet (Aguadé1996:198–199).3 Man findet aber auch andere Verbmodifikatoren wiezum Beispiel dà- in dem jüdischen Dialekt von Debdou (Pellat 1952:123, 133), qà-, südöstlich von Chefchaouen (Colin 1975:1195)), à-bei den Branes nördlich von Taza und in Ouargha (Colin 1921:98)oder ©à- bei den Írà©na (Guironnet 1954)).

In manchen prähilalischen Dialekten Nordmarokkos, zum Beispielin Anjra aber auch in anderen Gebieten von Jbala (¥bàla), kommtein Präverb là- vor (Vicente 2000:103–105).4 Dank der Untersuchungvon Vicente über den Dialekt von Anjra hat man sehr viele Beispielefür den Gebrauch dieses Verbmodifikators.5

1 Eine Ausnahme bildet der jüdische Dialekt von Sefrou; vgl. Stillman (1981:239;1988:40).

2 Einen Überblick über diese Verbmodifikatoren in Marokko sowie ihre Etymologienfindet man bei Aguadé (1996). Vgl. auch Durand (1991; 1994:89–90) und Grand’Henry(1978:215–220).

3 Dazu auch Colin in de Prémare (1993–1999: II, 4).4 Diese Arbeit ist besonders interessant, weil sie fast ausschliesslich auf den Aussagenweiblicher Informanten beruht. Vgl. auch de Prémare (1993–1999: XI, 3) und

(1986, Texte von Z>hra einer Frau aus dem Dorf Ó>mmàra).5 Der langsam auch von kà- verdrängt wird. Vicente (2000:103–104) bemerkt

dazu: “nos encontramos con una fuerte alternancia entre là- y kà-; siendo usado el primero mayoritariamente por mujeres de edad avanzada y personas jóvenesanalfabetas, y el segundo por niños y niñas escolarizados y por la mayoría de loshombres”.


14

Besonders interessant ist der Dialekt von Chefchaouen (”6f“àw6n),weil in dieser Mundart der Verbmodifikator flektiert wird (Natividad1998:116–117). Wenn zum Beispiel das Verb m“a ‘gehen’ zusammenmit dem Verbmodifikator konjugiert wird, so ergibt sich:

Sg. Pl.

3.m. là-yìm“i 3.c. là-yìm“ìw3.f. dà-≈6m“i,

là-≈6m“i2.c. dà-≈6m“i 2.c. dà-≈6m“ìw1.c. nà-n6m“i 1.c. nà-n6m“ìw

Es gibt inzwischen eine beträchtliche Anzahl von Beispielen für denGebrauch dieser Verbmodifikatoren in Chefchaouen. In einem Aufsatzmit Texten aus dieser Stadt, die von Natividad & Rahmouni (1996)ediert wurden,6 findet man folgende Sätze: nà-n>""ìw †-†˙ìn ‘wir tunMehl hinein’, là-y≥ìbu s6††à“ d-6l-b>y†à∆ ‘sie bringen sechzehn Eier’, madà-d6fh>m-“i ‘sie versteht nichts’, fìwàx là-≈6n'>s yìmm6ç ‘wann schläftdeine Mutter?’, 'là“, à-yìmma, dà-≈6xd>' hàd l>-'yàl ‘o Mutter, warumtäuschst du diese Kinder?’, là-y>'†ìha dàç5l ‘er gibt ihr etwas zu essen’.Und aus einer noch unveröffentlichten Doktorarbeit von Moscosoüber den Dialekt von Chefchaouen stammen folgende Beispiele:7

h5wwa hna là-yìsk5n ‘er wohnt hier’, ˙>tta dà-d5wß>l n-6d-dà¶ d-¶¶à≥6l‘bis sie das Haus des Mannes erreicht’, àna nà-n>h∂>¶ ‘ich sprechegerade’.

Der Verbmodifikator là- wird also in Chefchaouen folgendermas-sen flektiert:

Sg. Pl.

3.m. là- 3.c. là-3.f. dà-, là-82.c. dà- 2.c. dà-1.c. nà- 1.c. nà-

6 Natividad & Rahmouni (1996). Die von den Verfasserinnen verwendeteTranskription wurde hier von mir geringfügig geändert.

7 Moscoso, El dialecto árabe de Chauen. Moscoso weist darauf hin, dass die Flexiondes Verbmodifikators charakteristisch für die Sprache der Frauen ist. Männer gebrau-chen immer nur die unveränderliche Form là- oder sogar das gemeinmarokkani-sche kà-.

8 dà- und là- werden frei variiert.


LÀ- 15

Schon M. Cohen und Colin (1935:134) haben vor vielen Jahren dieVermutung geäussert, dass der nordmarokkanische Verbmodifikatorlà- aus dem Berberischen illa ‘er ist, es gibt’ entstanden sein könnte.9

Dieser Etymologie folgend, wies ich in einem 1996 erschienen Aufsatzdarauf hin, dass es noch dazu einen auffälligen Parallelismus zwi-schen den flektierten Formen aus Chefchaouen und der Konjugationdes Verbs ‘sein’ im Ta“el˙it gäbe, wie aus der folgenden Tabelleersichtlich wird:

Sg. Chef. Ta“el˙it Pl. Chef. Ta“el˙it

3.m. là- illa ‘er ist’ 3.c. là- llan ‘sie sind’3.f. dà- t6lla ‘sie ist’ (f. llant)2.c. dà- t6llit ‘du bist’ 2.c. dà- t6llam ‘ihr seid’

f. t6llamt)1.c. nà- lli© ‘ich bin’ 1.c. nà- n6lla ‘wir sind’

Da die Formen t6lla/t6llam > dà- als Sonorisierung t > d 10 und die1.c. Sg. nà- als Analogiebildung zur 1.c. Pl. leicht zu erklären wären,sprach ich mich ebenfalls für die Vermutung berberischer Substratein-wirkung in diesem Fall aus, freilich ohne einen arabischen Ursprungfür diesen Vermodifikator ganz auszuschliessen (Aguadé 1996:205–206).Nun, inzwischen halte ich diese Erklärung für ziemlich unwahrschein-lich, denn es gibt sonst gar keine Belege dafür, dass arabische Dialektein Marokko bei Entlehnungen aus dem Berberischen auch die ber-berische Flexion übernehmen.

Ausserdem gibt es eine viel einfachere Erklärung für die flektiertenFormen aus Chefchaouen: man kann sie nämlich ohne weiteres alsErgebnis einer Assimilation des Verbmodifikators là- an die nachfol-genden Präfixe der 2. und 3. Personen des Imperfekts deuten, also:

2. Sg. und Pl.: là- + d(6) . . . > dà-d(6) . . .3. Sg. und Pl.: là- + n(6) . . . > nà-n(6) . . .

Somit wären diese flektierten Verbmodifikatoren aus Chefchaouenlediglich Allomorphe des Präverbs là-: die Frage, ob dieser là- aus

9 Für Colin ist der Verbmodifikator là- “un résidu du berbère illà, 3è pers. masc.sing. du verbe d’existence”. Auch Lévy (1990:550) und Tilmatine (1999) halten dieseEtymologie für gesichert.

10 In diesem Dialekt ist die Sonorisierung des Imperfektpräfixes t- ohnehin dieRegel: tàk5l > dàç5l ‘sie wird essen’, t6bdàw > d6bdàw ‘ihr werdet beginnen’, t6m“i> d6m“i ‘du wirst gehen’ (Natividad & Rahmouni 1996:113–116).


dem Berberischen illa stammt muss allerdings noch diskutiert werden.

Für eine berberische Etymologie spricht schon manches. Wie bereitsvorhin erwähnt wurde, in nordmarokkanischen arabischen Dialekten—es handelt sich dabei ausschliesslich um prähilalische Mundarten—kommen die Verbmodifikatoren là-, dà-, à- und qà- vor. In manchenberberischen Dialekten (im Rif und Mittlerem Atlas) werden diePräverben la-, da-, a- und qa- verwendet, und zwar auch um dieaktuelle Gegenwart auszudrücken (Chaker 1984:972; Durand 1998:125;Kossmann 2000:121–122)—der Parallelismus zwischen arabischenund berberischen Dialekten ist in diesem Punkt auffällig. Die Subs-trateinwirkung scheint hier also eindeutig, zumal die betreffenden ber-berischen und arabischen Dialekte benachbart sind.11 Man bedenkeausserdem, dass einen berberischen Ursprung bei diesen Präverbenzu vermuten ja naheliegend ist—schliesslich hat man es im NordenMarokkos mit Dialekten zu tun, welche ohnehin einen starken Einflussdes berberischen Substrats aufweisen.12

Nun, streng genommen darf man allerdings nur dann berberischeSubstratwirkung annehmen, wenn die betreffende Erscheinung nichtan einer anderen Stelle des arabischen Sprachgebiets vorkommt, dasdieses Substrat nicht aufweist.13 Und das ist eben hier nicht der Fall,denn sowohl im Jemen als auch in Syrien—wo es bestimmt keinberberisches Substrat gegeben hat—kommt ein Verbmodifikator là-vor. Behnstedt (1985: Karte 83, Punkte 134, 140) erwähnt nämlichin seinem Atlas der nordjemenitischen Dialekte einen Vermodifikator

11 Auch Tilmatine (1999:107–108) weist auf einen möglichen berberischen Ursprungfür die marokkanischen Verbmodifikatoren là- und qà- hin.

12 Eine übersichtliche Zusammenfassung dieser Merkmale findet man bei Diem(1979:52–55). Neben manchen Erscheinungen, welche typisch für alle maghrebini-schen Dialekte sind (etwa Reduktion der kurzen Vokale und Silbenbildungsregeln),werden in den nordmarokkanischen Mundarten vor allem folgende durch Substra-teinwirkung erklärt: Affrizierung von t, d, b; ∂ > †; ma ‘Wasser’ wird als Pluralbehandelt; Veränderungen des syntaktischen Geschlechts einzelner Wörter; Genitiv-konstruktionen des Typs xtu dd6-m˙6mmed ‘die Schwester Mu˙ammads’. In der letz-ten Zeit hat sich Tilmatine (1999) mit dem berberischem Substrat in Algerien undMarokko befasst und eine ausführliche Aufstellung möglicher Fälle von Substratwirkungzusammengestellt. Zum Thema vgl. auch Aguadé & Vicente (1997:225–227, 234–236),sowie Lévy (1996).

13 Diem (1979:16): “Die betreffende arabische Erscheinung darf nicht an eineranderen Stelle des arabischen Sprachgebiets erscheinen, das dieses oder ein entspre-chendes anderes Substrat nicht aufweist, sofern die Verbreitung der Erscheinungdurch Migration oder Wellenbewegung ausgeschlossen ist”.

16


là- für die aktuelle und habituelle Handlung, der in der Gegend von'A††ah, Marays und Juban gebraucht wird.14 Vanhove (1993:191) hatlà- auch in der Ortschaft Dhàla' (südlich der von Behnstedt unter-suchten Dörfer) belegt. Als Beispiel führt sie an: là-ybki ‘er weintgerade’. Zur Etymologie dieses Präverbs werden von den genanntenAutoren keine Angaben gemacht.

Was Syrien betrifft, so erwähnt Behnstedt (1997:325) in seinemSprachatlas den Gebrauch eines Verbmodifikators là- um die Zukunftauszudrücken für die Gegend von Soukhne sowie für andere Ort-schaften.15 Es sei hier noch hinzugefügt, dass es auch in anderenorientalischen Dialekten Verbmodifikatoren gibt, welche eine durch-aus plausible (und rein arabische) Etymologie für den nordmarokka-nischen là- liefern können. Um die Zukunft auszudrücken gebrauchtman in manchen ägyptischen Dialekten (im Delta sowie in Oberägyptenum Gina, Luxor und Idfu) einen Verbmodifikator là˙ (Behnstedt &Woidich 1985: Karten Nr. 224, 225), welcher wohl aus ràyi˙ ‘gehend’entstanden ist.

Der nordmarokkanische Verbmodifikator là- kann genau so gutaus rà˙ (rà˙ > là˙ > là-)16 wie in den oben erwähnten ägyptischenDialekten entstanden sein. Und man sollte nicht ausser acht lassen,dass es im Hassaniyya—im Süden Marokkos und in Mauretanien—auch einen Verbmodifikator là- (für das Futur) gibt, dessen Ursprungman ebenfalls mit einer rein arabischen Etymologie (nämlich ausdem Partizip làhi ) erklärt (Cohen 1924:274; 1963:113).

Vanhove hat in ihrem oben erwähnten Aufsatz auf das Vorkommeneines Verbmodifikators là- in zwei so weit auseinanderliegendenLändern wie Jemen und Marokko hingewiesen und mit Recht dieberberische Etymologie für die nordafrikanische Partikel aus diesemGrund in Frage gestellt: “à la lumière des données yéménites (pourl’instant inexpliquées), l’origine berbère du préverbe marocain estpeut-être à discuter, notamment en raison de l’éloignement géo-graphique des zones berbérophones et arabophones concernées, à

14 In anderen Gegenden findet man là- und là'- um die Zukunft auszudrücken(vgl. Behnstedt 1985: Karte 8). Einen Verbmodifikator là findet man auch im neu-aramäischen Dialekt der Juden von Rustaqa (vgl. Khan 2002:402), der aber damitnichts zu tun hat.

15 Vgl. auch Behnstedt (1994:61).16 Das Verb rà˙ ist in Marokko auch gebräuchlich, hat aber meistens die Bedeutung

‘[am Abend] zurückkehren, nach Hause zurückkehren’: vgl. de Prémare (1993–1999:V, 231) und Harrell (1966:122).

LÀ- 17


moins que l’histoire ne montre des liens privilégiés entre elles”(Vanhove 1993:191 und Anm. 16).

Nun, Migrationen hat es schon gegeben aber immer vom Jemennach Nordafrika und der iberischen Halbinsel so dass manche jeme-nitische Züge sich sogar im andalusischen Dialekt nachweisen lassen(Corriente 1989): diese Einflüsse und die Tatsache, dass man là- auchin Syrien und Ägypten findet spricht eindeutig gegen eine berberi-sche Etymologie für den nordmarokkanischen Verbmodifikator.

Bibliographische Hinweise

Aguadé, Jordi. 1996. “Notas acerca de los preverbios del imperfectivo en árabe dia-lectal magrebí”. Estudios de dialectología norteafricana y andalusí 1.197–213.

——, & Ángeles Vicente. 1997. “Un calco semántico del bereber en árabe dialec-tal magrebí: El uso de la preposición 'la en el comparativo”. Estudios de dialectologíanorteafricana y andalusí 2.225–240.

Behnstedt, Peter. 1985. Die nordjemenitischen Dialekte. I. Atlas. Wiesbaden: L. Reichert.—— 1994. Der arabische Dialekt von Soukhne (Syrien). II. Phonologie, Morphologie Syntax.

III. Glossar. Wiesbaden: O. Harrassowitz.—— 1997. Sprachatlas von Syrien. Kartenband. Wiesbaden: O. Harrassowitz.——, & Manfred Woidich. 1985. Die ägyptisch-arabischen Dialekte. II. Dialektatlas von

Ägypten. Wiesbaden: L. Reichert.Chaker, Salem. 1984. “Aspect”. Encyclopédie Berbère VII, 971–977.Cohen, David. 1963. Le dialecte arabe ˙assànìya de Mauritanie (parler de la G6bla). Paris:

Klincksieck.Cohen, Marcel. 1924. Le système verbal sémitique et l’expression du temps. Paris: Imprimerie

Nationale.Colin, Georges S. 1921. “Notes sur le parler arabe du nord de la région de Taza”.

Bulletin de l’Institut Français d’Archéologie Orientale 18.33–119.—— 1935. “L’opposition du réel et de l’éventuel en arabe marocain”. Bulletin de

la Societé de Linguistique de Paris 36.133–140.—— 1975. “al-Mahgrib, 7: Aperçu linguistique”. Encyclopédie de l’Islam (Nouvelle Édi-

tion) V, 1193–1198.Corriente, Federico. 1989. “South Arabian features in Andalusi Arabic”. Studia lin-

guistica et orientalia memoriae Haim Blanc dedicata, ed. by Paul Wexler, AlexanderBorg & Sasson Somekh, 94–103. Wiesbaden: O. Harrassowitz.

Diem, Werner. 1979. “Studien zur Frage des Substrats im Arabischen”. Der Islam56.12–80.

Durand, Olivier. 1991. “I preverbi dell’imperfettivo in arabo dialettale”. Rivista degliStudi Orientali 65.1–10.

—— 1994. Profilo di arabo marocchino: Varietà urbane centromeridionali. Rom: Universitàdegli Studi La Sapienza.

—— 1998. Lineamenti di lingua berbera: Varietà tamazight del Marocco centrale. Rom:Università degli Studi La sapienza.

Grand’Henry, Jacques. 1976–1978. “La syntaxe du verbe en arabe parlé maghrébin”.Le Muséon 89.457–475; 90.237–258; 439–456; 91.211–224.

Guironnet, M 1954. “©à- un outil linguistique polyvalent chez les Ulàd Ya'qub desSrà©na”. Hespéris 41.425–433.

18


Harrell, Richard. S. 1966. A Dictionary of Moroccan Arabic: Moroccan-English. Washington,D.C.: Georgetown University Press.

Khan, Geoffrey. 2002. “The Neo-Aramaic dialect of the Jews of Rustaqa”. ‘Sprichdoch mit deinen Knechten aramäisch, wir verstehen es!’: 60. Beiträge zur Semitistik. Festschriftfür Otto Jastrow zum 60. Geburtstag, hrsg. von Werner Arnold & Hartmut Bobzin,395–409. Wiesbaden: O. Harrassowitz.

Kossmann, Maarten. 2000. Esquisse grammaticale du rifain oriental. Paris & Louvain:Editions Peeters.

Lévy, S. 1990. Parlers arabes des Juifs du Maroc: Particularités et emprunts. Thése deDoctorat d’État, Université Paris VIII.

——. 1996. “Repères pour une histoire linguistique du Maroc”. Estudios de dialectologíanorteafricana y andalusí 1.127–137.

Natividad, Emma. 1998. “Le dialecte de Chefchaouen”. Peuplement et arabisation auMaghreb occidental: Dialectologie et histoire, hrsg. von Jordi Aguadé, Patrice Cressier& Ángeles Vicente, 109–120. Madrid: Casa de Velázquez & Zaragoza: Universidadde Zaragoza.

——, & Aicha Rahmouni. 1996. “Textos árabes de Chefchaouen: Transcripción,traducción y glosario”. Estudios de dialectología norteafricana y andalusí 1.139–155.

Pellat, Charles.1952. “Nemrod et Abraham dans le parler arabe des Juifs de Debdou”.Hespéris 39.121–129.

Prémare, Alfred-Louis. de. 1986. La tradition orale du Mejdûb: Récits et quatrains inédits.Aix-en-Provence: Edisud.

—— 1993–1999. Dictionnaire arabe français. 12 Bde. Paris: L’Harmattan.Stillman, Norman A. 1981. “Some notes on the Judeo-Arabic dialect of Sefrou”.

Studies in Judaism and Islam presented to Shelomo Dov Goitein, hrsg. von Shelomo Morag,Issachar Ben-Ami & Norman A. Stillman, 230–251. Jerusalem: Hebrew University.

—— 1988. The language and culture of the Jews of Sefrou, Morocco: An ethnolinguistic study.Manchester: University of Manchester.

Tilmatine, Mohamed. 1999 “Substrat et convergences: Le berbère et l’arabe nord-africain”. Estudios de dialectología norteafricana y andalusí 4.99–119.

Vanhove, Martine. 1993. “Note sur le dialecte q6ltu de Dhala' (province de Lahej,Yémen)”. MAS-GELLAS (N.S.) 5.175–199.

Vicente, Ángeles. 2000. El dialecto árabe de Anjra (norte de Marruecos): Estudio lingüísticoy textos. Zaragoza: Universidad de Zaragoza.

LÀ- 19


This page intentionally left blank

21

VARIABILITY REPRODUCED: A VARIATIONIST VIEW OF THE [ π̨ ]/[Î] OPPOSITION

IN MODERN ARABIC DIALECTS*

Enam al-WerUniversity of Essex, UK

1. Introduction

In traditional and urban Arabic dialectology it is customary to referto the change from interdental sounds to stop sounds as a case ofphonological merger. Naturally, in the technical usage of the term‘merger’ it is assumed that, at some point in the distant or recentpast, an earlier state of affairs existed when the features in questionwere kept separate by the native speakers. Often we have evidencefor phonological mergers in neighbouring and/or related dialectswhich did not undergo these processes, or in the existence of relicforms which preserve the earlier distinction, or, where mergers are asyet incomplete, evidence can be found in intergenerational differences,and other social stratification patterns, which provide attestations inapparent time of the pre-merger state. In the case of the plain inter-dental Arabic sounds, the merger of [∆] and [≈] with [t] and [d],respectively, is historically and empirically substantiated. However,evidence that a phonological merger has occurred which collapsedthe lexical set containing modern [µ] with that containing modern[∂] (in either direction) is simply unavailable. None of the spoken

* This article is a revised version of a paper originally presented at AIDA 4 inMarrakech. I wish to thank Professor F. Corriente for his promptness in respond-ing to my electronic queries. Thanks also to the editors of this volume for the sug-gestions and insightful comments, many of which have been incorporated, whilesome have fallen on deaf ears. Any remaining pitfalls are entirely my own. Twoof my colleagues at Essex deserve mentioning: Mike Jones for being the scholar heis: inspiring and generous, and Andrew Spencer for the many hours of enjoyableand fruitful discussions. To both of them I am truly grateful.


22 -

dialects maintains the distinction presumed in the written form ofthe Standard variety, no spoken dialect has both sounds in its pho-netic inventory, and no dialect, however isolated, ‘preserves’ vestigialforms of the distinction. In this article, I debate this rather peculiarsituation, and I suggest an alternative analysis.

In relation to the occurrence of interdental sounds, the phoneticinventories of modern Arabic dialects can be classified into two types:

• Type I dialects, which represent a system empty of interdentalsounds altogether, as in the majority of the urban varieties in theLevant. These dialects are assumed to have undergone a systematicsound change, which has resulted in interdental and stop merger,as illustrated below:

Schema 1: Scenario of events in Type I dialects

Plain sounds Outcome Phonetic property of the outcome

/∆/, [∆] � [t] /t/ [t], contains lexical sets withetymological /∆/ and /t/.

/≈/, [≈] � [d] /d/ [d], contains lexical sets withetymological /≈/ and /d/.

Emphatic sounds

/µ/, [µ] � [∂] /∂/ [∂], contains lexical sets withetymological /µ/ and /∂/.

The course of events envisaged for the plain interdental and stopmergers as outlined in Schema 1 is empirically tenable, as can beevidenced in the dialects which did not undergo the merger, and,perhaps even more convincingly, in the dialects which show stablevariability in the use of plain interdental and stop, or progressivetendency towards the merger. But, empirical evidence for the mergerof the emphatic sounds (in this case in favour of the stop sound) isunavailable in Arabic dialects, since there are no native dialects whichmaintain the distinction. Theoretically, however, a sound changewhich would have shifted the emphatic interdental sound to a stopsound (to parallel the changes in the plain interdentals) is quite plau-sible, given that interdental sounds occur relatively rarely in humanlanguages, and a change from fricative interdental to stop is a widelyattested phenomenon. We can assume that the system here simplyeliminates ‘cumbersome’ sounds wholesale.


23

• Type II dialects, which have the full set of interdentals, plain [∆]and [≈], and emphatic [µ]. The traditional (indigenous) dialects ofJordan and many of the rural and Bedouin Levantine varieties areexamples of such dialects. In these dialects, the plain interdentalsare phonemically distinguished from their stop counterparts [t] and[d], respectively, but the emphatic interdental has no stop counter-part. Schematically:

Schema 2: Scenario of events in Type II dialects

Plain sounds

/∆/ [∆], /t/ [t] No merger/≈/ [≈], /d/ [d] No merger

Emphatic sounds outcome Phonetic property of the outcome

/∂/ � /µ/ /µ/ [µ], contains lexical sets withetymological /∂/ and /µ/, i.e. mergerin favour of the interdental.

Schema 2 appears problematic, in that the change is incredibly selec-tive (singles out one sound from a system), and predicts a develop-ment in the ‘wrong’ direction: from stop to interdental fricative indialects which continue to maintain a phonemic distinction betweenother stop and interdental sounds. A change from stop to fricativeis, of course, attested in human languages, but such a developmenttends to be systematic, as exemplified by the systematic shift of Proto-Indo-European voiceless stops to fricatives (Grimm’s Law). OriginalArabic ∂àd, however, was not a stop sound, but most probably anemphatic lateral fricative, which lost lateralisation at some stage. Thisissue and its implications will be further explored later in this article,but for now let us start from the premise that in both cases abovethe emphatic sounds had undergone different and/or additional his-torical developments from those which had affected the plain sounds.As I will show presently, these differences are reflected in the pat-terns of variation and behaviour of the plain versus the emphaticsounds in contemporary Arabic dialects.


2. Data and discussion

In 1987, I investigated variation and change in the use of the inter-dental sounds in three Jordanian towns, and on the basis of thisresearch I reported that these sounds were undergoing a linguisticchange which merged them with their stop counterparts (details arein Al-Wer 1991). I, like other researchers, who investigated thesevariables in Jordan or in neighbouring communities, treated thechange as a systematic development. A closer look at the statisticaldetails in my data and those provided by other researchers revealsthat the plain interdentals show a slower rate of change to stopsthan the emphatic interdental change to emphatic stop (see Abdel-Jawad & Awwad [1989] on a sample of Jordanian speakers, Al-Khatib[1988] on the city of Irbid in north Jordan, and Jassem [1993] ona sample of refugees from the Golan Heights now resident inDamascus; to my knowledge these are the only studies to date whichhave included the interdentals as variables). In these studies, thechange (if these are indeed cases of change in progress) from [µ] to[∂] is considerably faster than the changes from [∆] to [t] and [≈]to [d]. The plain interdentals [∆] and [≈] show very similar rates. Irevisited the town of Sult in Jordan (one of the towns investigatedin 1987) ten years later to collect data for a separate research pro-ject. The smaller sample of data I have from the 1997 study pro-vides information in real time on the progression of the change frominterdental to stop. The data show that the change has progresseddramatically; in particular, the change from [µ] to [∂] now showsnear completion in the speech of young women, while the frequencyof occurrence of /∆/: [t] and /≈/: [d] has roughly doubled (from28% to 45%). In my opinion, the differences in the behaviour ofthe plain sounds and the emphatic sounds may not be accidental.Rather, they suggest that the parameters governing the changes from[∆] and [≈] to [t] and [d] are different from the parameters gov-erning the change from [µ] to [∂], which may be related to the factthat the change which affects the plain interdentals is a merger (aphonological change), but the change which affects the emphatic inter-dental is a straightforward sound change from fricative to stop (aphonetic change). In turn, this raises the question of the member-ship of the emphatic interdental sound in this set, and implies thatplain and emphatic interdentals should be analysed separately (notas a system).

24 -


It is not clear whether there is any basis for expecting mergers,as a type of sound change, to progress differently from regular soundchanges which do not involve phonological mergers, which would,for instance, explain the varying rates of change presented above.On the basis of empirical data, variationist sociolinguists have devel-oped a number of models which show that there are several radi-cally different mechanisms of merger, which operate at different rates(see for instance Trudgill & Foxcroft 1978; Labov 1994, chapter 10).The case of the plain interdental and stop merger in Arabic dialectsfits in with the model referred to by Trudgill and Foxcroft (1978)as ‘merger by transfer’, which involves a gradual transfer of wordsfrom one category to another. Labov (1994:323) maintains that mergerby transfer is the slowest, compared with ‘merger by approximation’and ‘merger by expansion’.

Sociolinguistic stereotyping of the interdental sounds in Jordan pro-vides further indication that the plain and emphatic sounds shouldbe treated separately. One observes, for instance, that [µ] is extremelystigmatised in Jordan (and in the Levant in general). This sound isoften used to mimic, in ridicule, speakers of the dialects which haveit, but no such stereotypes are associated with the use of [∆] or [≈].The issue of how and why positive or negative social values areattached to certain sounds is complex, and the explanations pro-posed are often circular. It is possible, however, that when a soundchange is nearing completion and the use of the departing feature,here [µ], becomes restricted to a small minority of speakers, itsmarkedness increases, and the sound itself is then perceived as obso-lete and rustic.

An intriguing case of plain and emphatic interdentals progressingdifferently, and in this case dramatically so, is reported in de Jong’s(2000) description of the Biyyàµiy dialect in northern Sinai. In thisdialect the plain interdentals have shifted to stop but the emphaticinterdental [µ] largely remains unaffected, a further indication that,in the mechanisms of change, while the plain interdentals move sys-tematically, the emphatic interdental does not necessarily follow suit.It appears that in Biyyàµiy /µ/ is not involved in variation, apartfrom the cases where [∂] is used invariably in K-forms (de Jong2000:331–332). The stability of [µ] in comparison with [∆] and [≈]in this dialect is itself an interesting, but, I think, different issue. DeJong (2000:331–332, n. 8) cites examples of other Arabic dialects“. . . in which interdentals have developed less predictably . . .”, namely

25


the dialects of Bahrain, Siirt, Àz6x, and B6˙zàni. I understand ‘lesspredictably’ here to mean that the interdentals [∆], [≈], [µ] in thesedialects did not shift to stop counterparts, but to labiodental/stop/stopin Ba˙àrna Arabic (i.e. the dialect of Bahrain), labiodental/labioden-tal/labiodental in Siirt, and sibilant/sibilant/sibilant in Àz6x andB6˙zàni. Less predictable these cases may be, but the developmentsare perfectly in line with principles of simplification through elimi-nation of marked sounds in favour of more frequent and/or lessmarked sounds. In comparison however, the Biyyàµiyya case is baffling.For no apparent linguistic benefit, it disregards the time-honouredconstraint of economy in language change. De Jong’s proposal thatexternal social factors such as group identity may account for theanomaly of the Biyyàµiyya system is plausible, especially when oneconsiders that the name of the tribe itself contains the emphaticinterdental sound. Take the following example for an analogy. InAmman, the sound [g] has largely lost ground to ["], but traditionalJordanian names are immune to this transition. Speakers who other-wise consistently use ["] maintain [g] in proper names such as 'uglaand 'gàb.

A phenomenon which is readily observable in the Arabic dialectswhich have the interdental sounds (Type II dialects) is that the speak-ers of these dialects very often read the orthographic representationof ∂àd as [µ], while the speakers whose native dialects do not includethe interdentals read the interdentals as stops. Take, for instance,the following sets of data as examples of the performance of someArabic speakers in reading tasks, a domain in which speakers usu-ally monitor their speech quite closely (not the sort of data one nor-mally seeks, I dare say). Standard Arabic norms are highly prescribed,and given that newsreaders, in particular, are expected to adhere tothese norms, the difficulties they exhibit with the interdental soundsare probably ‘beyond their control’, so to speak.

In the fist set I recorded four Moroccan speakers, from the Moroccansatellite TV station. Two of the speakers were delivering formalspeeches from a written text, and two newsreaders. The speakers inthis set read all instances of interdental sounds, both plain andemphatic, as stops. A sample is listed in Table 1 below (the soundsin question are in square brackets).

26 -


Table 1: Examples from the Moroccan speakers

Correct reading Actual reading English gloss

li [≈] àlika li [d] àlika Hence[∆] alà [∆] at 'a“ar [t] alà [t] at 'a“ar Thirteen[∆] àbit [t] àbit Steady, Stable"a [∆∆] àniya "a [tt] àniya The second"ittixà [≈] "ittixà [d] The undertaking (of a task)muna [µµ] amàt muna [∂∂] amàt Organisationsyanta [µ] ir yanta [∂] ir He is awaiting

The Moroccan ‘errors’ are readily explainable with reference to theabsence of interdental sounds from the native dialects of the speak-ers (assuming they are speakers of such dialects). The behaviour ofthese speakers is, in essence, similar to the behaviour of, say, Germanlearners of English. In another set of recordings I made, the speakerscame from Jordan, Tunisia, Iraq and Yemen (presumably speakersof Type II dialects, which preserve the interdental sounds). Twospeakers of each locality were recorded from national satellite TVstations, delivering formal speeches from written texts, or reading thenews bulletin. Here, the speakers make no ‘mistakes’ at all with theplain interdentals, but they read ∂àd as (voiced emphatic interden-tal) [µ]. This error was most consistent in the case of the Tunisianspeakers, who read all instances of /∂/ as [µ], followed by theYemenis’ score of 94% wrong, Iraqi’s 76%, and the Jordanians’ 37%.A sample of the Tunisian data is listed in Table 2 below.

Table 2: A sample of the Tunisian reading of ∂àd

Correct reading Actual reading English gloss

mufàwa [∂] àt mufàwa [µ] àt Negotiations"alwa [∂] ' "alwa [µ] ' The situation"ay [∂] an "ay [µ] an Also[∂] idd [µ] idd Against"i [∂] àfatan "i [µ] àfatan Besides, in addition tofa [∂] lan fa [µ] lan Apart from (the fact that . . .)qa [∂] iyya qa [µ] iyya Issue

27


28 -

If a phonemic distinction between [∂] and [µ] ever existed, thesespeakers clearly have no cognitive access to such a distinction (whereasthey have no problem in making the distinction between plain inter-dental and plain stop). The data from the Jordanian speakers is par-ticularly interesting. The modern urban dialects in Jordan are currentlyin a state of variability with respect to the interdental sounds: [∆] isused alongside [t], [≈] is used alongside [d], and [µ] (for historical/µ/ and /∂/) is used alongside [∂] (for historical /µ/ and /∂/). Theattested variability represents changes in progress, i.e. these dialects(originally Type II) are undergoing a change from interdental to stopfor both plain and emphatic sounds (i.e. are becoming Type I). But,it is noticeable that the speakers have no problems in re-splitting /t/into [t] and [∆], and /d/ into [d] and [≈]. The difficulty they showis confined to the distinction between the emphatic variants. I attributethis difference to the fact that in the case of the plain consonants,the phonemic distinction is available in the linguistic experience ofthe speakers, even if, sometimes, they do not make this distinctionphonetically, whereas in the case of the emphatic consonants, thereis no such evidence on which speakers can make a phonemic splitwhen required, e.g. speaking or even reading the Standard variety.

3. ∂àd versus µà" in history

If one examines the historical developments of /∂/ and /µ/ andtheir distribution in the old Arabic dialects, the existence of a stagewhere [∂] and [µ] were distinguished phonemically becomes precar-ious. The thesis that the emphatic stop descends from Proto-Semiticvoiced lateral fricative, say [�], is now widely accepted. To my mind,Steiner (1976) and Corriente (1974) present a strong argument inthis direction. Corriente (1974:51), starting from the stage where theemphatic voiced lateral fricative was phonemically distinguished from/µ/, outlines a convincing course of events leading to the presentsituation. These developments can be envisaged as follows:

• The system contains /µ/ [µ], and /�/ [�]. Given the phoneticcomplexity associated with the production (and the acquisition) ofthe lateral fricative, it is reasonable to suggest that it lost its markedphonetic properties, and changed into something else. The ques-tion is: what could it have changed into?. The simple answer is


that it would have changed into the nearest, less marked, andacoustically similar sound. If it were to change into a stop by los-ing lateralisation and creating a total closure, important acousticinformation, namely friction, would have been lost. At this point,it is worth noting that the complexity of the sound is much morelikely to cause problems for the learners (the new acquirers, thechildren) than for the adults, and therefore acoustic clues are vitalfor a successful reproduction. The most likely development, there-fore, was that the lateral fricative lost lateralisation, and the fric-tion was reproduced through an interdental narrowing. This meansthat its phonetic (and acoustic) properties became very similar toor identical with [µ], and the outcome was a merger in favour of[µ], which then contained the lexical sets with etymological lateralfricative as well as those which already had the emphatic inter-dental. Type II dialects stopped here.

But, Type I dialects continue as follows:

• [µ] (which is now the result of the merger of [�] and [µ]), andthe plain interdentals [∆] and [≈] change into stops [∂], [t] and[d], respectively. Notice that the outcome of this change is a mergerin the case of the plain interdentals (since the system already contained/t/ and /d/), but the outcome in the case of the emphatic soundis not a merger, since the system did not have a voiced stopemphatic sound; the latter is rather a straightforward sound changewith no mergers involved.

What the historical developments outlined here show is that TypeII dialects have never had a stop counterpart to [µ] in the first place,and that neither Type I nor Type II dialects have ever had the [µ]versus [∂] opposition. If an opposition never existed, then the soundchange, to [∂], which has affected or is affecting [µ] in many Arabicdialects is not a case of merger. Strictly speaking, however, in thecase of Type I dialects, the change from interdental to stop musthave gone through a stage of variability, whereby the old and thenew forms were used. These dialects may have contained [µ] and[∂] variably at some stage, as variants (allophones) of the samephoneme, but not contrastively, i.e. a very similar situation to thatfound in some contemporary Jordanian dialects, as explained above.

29


30 -

4. The phonemic status of /∂/ versus /µ/ in Classical Arabic

Written Classical Arabic suggests a phonemic distinction between/∂/ and /µ/, and at first sight, the phonetic descriptions by theancient grammarians are suggestive of the existence of this distinc-tion. Nevertheless, it is surprising from this viewpoint that none ofthe spoken dialects has maintained this contrast. The comments bythe ancient grammarians are unclear, in that although they mentiondifferent pronunciations, for instance Sìbawayhi described a soundwhich in modern terms is likely to be a voiced lateral fricative [�],and a separate sound [µ], it is possible, in my opinion, that whatthe ancient grammarians are describing is in fact a situation of vari-ation similar to that observed in the modern dialects, where bothsounds occur but not in the same dialects. Moreover, it is clear thatthe modern dialects did not evolve directly from the Classical stan-dard, in so far as standard forms of languages are typically a moreor less conscious amalgam of dialects. The question then is whetherthe dialects from which both Classical Arabic and the colloquial vari-eties developed had this contrast. Indeed, in the pre-Islamic texts,different dialects had different variants, and there are numerousexamples of the alternation of the two sounds in the same words inthe early Islamic era (see El-Gindi 1982). El-Gindi (1982:426–428)also names at least seven different collections of the Qur"àn (maßà˙if )which either use ∂àd or µà" to represent both sounds throughout.Interestingly, he mentions that in the Qur"àn, the word [©ali:µ] is aperfect rhyme of ['ari:∂]. Clearly, these words can only rhyme if ∂àdand µà" had an identical phonetic realisation.

A possible conclusion from these observations is that the contrastfound in Classical Arabic between [µ] and [∂] represents an artificialsplit, which may have come about in a variety of ways, for instance,through an attempt to codify this variation on a lexical basis withthe possibility that lexemes incorporated in the standard may havecome from different [∂] versus [µ] dialects. Those speakers in mydata who do not make the ‘correct’ distinctions in reading ClassicalArabic may not in fact be making mistakes, but are simply main-taining, or reproducing, the same variability which was there betweenthese sounds in the classical period. Finally, Clive Holes, who ishereby dearly thanked, has brought to my attention the followingcomment by Kofler (1940–1942:96–97), which sums up the situa-tion aptly:


31

“. . . ferner auch, dass bereits im Altarabischen kein

APPROACHES TO ARABIC DIALECTS - Internet Archiveon many aspects of Arabic dialects, from negative constructions to folk tales, from diglossia to phonology. His didactic materials for

Documents